INDEX

Explanations

specific structured phrases and legal terminology

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ilyn

-0.16

uania

-0.16

 Fallon

-0.15

anel

-0.15

ovol

-0.15

Pairs

-0.14

itz

-0.14

 Pair

-0.14

erot

-0.14

ci

-0.13

POSITIVE LOGITS

IRON

0.16

 gent

0.15

ÙĦÙħØ§ÙĨ

0.14

estring

0.14

ordes

0.14

upe

0.13

 doses

0.13

azzi

0.13

enty

0.13

phin

0.13

Activations Density 0.008%

specific structured phrases and legal terminology

No Comments

No Known Activations